The objectives of grade control remain unchanged but new methods are being developed to attain them. Since the early-1990s, an improvement in the power of mine site computers has made the geostatistical tools of conditional simulation more accessible. Optimisation of the grade control process on mine sites now involves two stages: optimisation of the sampling regime and optimisation of the interpretation of ore and waste boundaries. Optimisation of the sampling regime first requires an understanding of the nature of the mineralisation and in particular the impact of particle size on the sampling precision. This relationship, termed the Fundamental Error of sampling, enables an optimum subsampling protocol to be defined for the sample preparation and assaying. However, the actual sample collection method, whether drilling or trenching, must be correctly selected and the sampling grid should be optimised. The relative costs and benefits of various alternative sampling methods and grids can be determined using the method discussed. The use of the grade control data can also be optimised. In the past ore and waste boundaries were often defined using a linear interpolation algorithm that smoothed the data (such as polygonal, sectional, inverse distance squared or even kriging methods). Frequently the relative impact of ore loss and dilution could not be estimated. Conditional simulation to generate equiprobable alternative maps (realisations) can now be adopted to address these issues. A transfer function is then used to evaluate the similarities and differences between the maps. The Ore Block Optimisation (OBO) process is a transfer function that produces a three-dimensional model of the optimal ore block outlines. This provides a clear guide for the mining geologist to refine and mine. The nature of the transfer function and the inputs and outputs are discussed. In the past, reconciliation has been the `acid test' for any grade control system. Any ore blocking method based on conditional simulation can make use of the exhaustive data set to ensure that reconciliation between prediction and production will improve significantly. The OBO method provides a guide to the mining geologist to ensure that the best ore blocks are mined. Misclassification between ore and waste is reduced, ore loss and dilution are minimised (and are measurable) and ore is optimally assigned to stockpiles of various cut-off grades. These advantages significantly enhance mine profitability and control of the mined grade.